NDFT: Accelerating Density Functional Theory Calculations via Hardware/Software Co-Design on Near-Data Computing System
Qingcai Jiang 1 ∗ , Buxin Tu 1 ∗ , Xiaoyu Hao 1 , Junshi Chen 1 , 2 and Hong An 1 , 2 , † · 2025
Read NDFT: Accelerating Density Functional Theory Calculations via Hardware/Software Co-Design on Near-Data Computing System by Qingcai Jiang 1 ∗ , Buxin Tu 1 ∗ , Xiaoyu Hao 1 , Junshi Chen 1 , 2 and Hong An 1 , 2 , † (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-03451.txt.
Description
Qingcai Jiang and co-authors present a computing paper, published in 2025, speeding up a core method for predicting how materials respond to light. Density functional theory in its time dependent form forecasts excited state behavior, but its steps are bound by memory traffic more than raw arithmetic. Shuttling data to separate accelerator chips wastes much of the effort.
Their system, NDFT, instead computes near the data on a processor with built in near memory units. A scheduler sends each part of the calculation to the unit that suits it best, and one heavy kernel gets a joint hardware software tune up. On a large test system the design ran several times faster than both plain processor and graphics chip baselines.
Catalog record
| CREATOR | Qingcai Jiang 1 ∗ , Buxin Tu 1 ∗ , Xiaoyu Hao 1 , Junshi Chen 1 , 2 and Hong An 1 , 2 , † |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-03451